On 7/29/24 1:49 AM, gregkh@xxxxxxxxxxxxxxxxxxx wrote: > > The patch below does not apply to the 5.15-stable tree. > If someone wants it applied there, or to any other stable or longterm > tree, then please email the backport, including the original git commit > id to <stable@xxxxxxxxxxxxxxx>. > > To reproduce the conflict and resubmit, you may use the following commands: > > git fetch https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/ linux-5.15.y > git checkout FETCH_HEAD > git cherry-pick -x 943ad0b62e3c21f324c4884caa6cb4a871bca05c > # <resolve conflicts, build, test, etc.> > git commit -s > git send-email --to '<stable@xxxxxxxxxxxxxxx>' --in-reply-to '2024072940-parish-shirt-3e49@gregkh' --subject-prefix 'PATCH 5.15.y' HEAD^.. Here's the 5.10-stable and 5.15-stable variant of this. For some reason this wasn't queued for 5.10 (probably because of the long ago 5.15-stable backport of io_uring to 5.10), but it should go there too. -- Jens Axboe
From f416e4a185dc3d007ea478db3ff75f47cd689eb6 Mon Sep 17 00:00:00 2001 From: Pavel Begunkov <asml.silence@xxxxxxxxx> Date: Wed, 10 Jul 2024 18:58:18 +0100 Subject: [PATCH 1/2] kernel: rerun task_work while freezing in get_signal() commit 943ad0b62e3c21f324c4884caa6cb4a871bca05c upstream. io_uring can asynchronously add a task_work while the task is getting freezed. TIF_NOTIFY_SIGNAL will prevent the task from sleeping in do_freezer_trap(), and since the get_signal()'s relock loop doesn't retry task_work, the task will spin there not being able to sleep until the freezing is cancelled / the task is killed / etc. Run task_works in the freezer path. Keep the patch small and simple so it can be easily back ported, but we might need to do some cleaning after and look if there are other places with similar problems. Cc: stable@xxxxxxxxxxxxxxx Link: https://github.com/systemd/systemd/issues/33626 Fixes: 12db8b690010c ("entry: Add support for TIF_NOTIFY_SIGNAL") Reported-by: Julian Orth <ju.orth@xxxxxxxxx> Acked-by: Oleg Nesterov <oleg@xxxxxxxxxx> Acked-by: Tejun Heo <tj@xxxxxxxxxx> Signed-off-by: Pavel Begunkov <asml.silence@xxxxxxxxx> Link: https://lore.kernel.org/r/89ed3a52933370deaaf61a0a620a6ac91f1e754d.1720634146.git.asml.silence@xxxxxxxxx Signed-off-by: Jens Axboe <axboe@xxxxxxxxx> --- include/linux/sched/signal.h | 6 ++++++ kernel/signal.c | 8 ++++++++ 2 files changed, 14 insertions(+) diff --git a/include/linux/sched/signal.h b/include/linux/sched/signal.h index 9743f7d173a0..13c95782c063 100644 --- a/include/linux/sched/signal.h +++ b/include/linux/sched/signal.h @@ -347,6 +347,12 @@ extern void sigqueue_free(struct sigqueue *); extern int send_sigqueue(struct sigqueue *, struct pid *, enum pid_type); extern int do_sigaction(int, struct k_sigaction *, struct k_sigaction *); +static inline void clear_notify_signal(void) +{ + clear_thread_flag(TIF_NOTIFY_SIGNAL); + smp_mb__after_atomic(); +} + static inline int restart_syscall(void) { set_tsk_thread_flag(current, TIF_SIGPENDING); diff --git a/kernel/signal.c b/kernel/signal.c index c7dbb19219b9..08bccdbb1b46 100644 --- a/kernel/signal.c +++ b/kernel/signal.c @@ -2579,6 +2579,14 @@ static void do_freezer_trap(void) spin_unlock_irq(¤t->sighand->siglock); cgroup_enter_frozen(); freezable_schedule(); + + /* + * We could've been woken by task_work, run it to clear + * TIF_NOTIFY_SIGNAL. The caller will retry if necessary. + */ + clear_notify_signal(); + if (unlikely(READ_ONCE(current->task_works))) + task_work_run(); } static int ptrace_signal(int signr, kernel_siginfo_t *info) -- 2.43.0
